12-Oct-85 17:47:22-MDT,00000003019;000000000001 Date: 12 Oct 1985 1747-MDT From: Beebe (Nelson H.F. Beebe) Subject: [Beebe (Nelson H.F. Beebe): WordStar to LaTeX translation] To: op.hendry, fh.harris cc: BEEBE 2 --------------- Date: 11 Oct 1985 1804-MDT From: Beebe (Nelson H.F. Beebe) Subject: WordStar to LaTeX translation To: C-GRISS, CH.JM.DOWNING, FH.HARRIS, MA.PETERSEN, OP.BOWMAN, OP.DUNCAN, To: OP.HENDRY, OP.LICHTY, OP.WILLETT, VI.FULLERTON, GU.KILLIAN cc: BEEBE, MA.HOPPENSTEADT, CH.JM.KLING, BI.PARKINSON, bi.kofOID, CH.MAYNE, cc: vi.leaver, ma.othmer I completed today a pair of programs for assisting in the conversion of files from WordStar (a popular microcomputer word processor) to LaTeX. The subject manuscript was Pete Conwell's Ph.D. thesis, which Ed Leaver is going to massage further into a form that LaTeX will accept. WordStar file format makes extensive use of the high-order bit of 8-bit ASCII characters to flag "hard" and "soft" line breaks, underlining, subscripting, superscripting, boldfacing, etc. The exact sequences to do this are described in a document which the authors of WordStar market for $500; needless to say, I did not buy it, but just reverse engineered the job. The WordStar file was KERMITed to the DEC-20 from an IBM PC by issuing SET FILE 8-BIT then SERVER to KERMIT-20, followed by a normal KERMIT-PC SEND command. Both programs are in the directory APS:. The first of them, WS87.PAS, can be used to copy any 8-bit file to a 7-bit file, turning all characters in the range 128 .. 255 into printable \nnn octal escape sequences. This allows examination of the WordStar file on the DEC-20, and could be used for other 8-bit files as well. The second, WSLTEX, is a 535-line workhorse which identifies all the WordStar magic sequences used in Pete's thesis and turns them into appropriate LaTeX commands. It would be straightforward to modify WSLTEX now for any other document formatter, and if necessary, to recognize other WordStar sequences not used in the thesis. Conversion time was 19 sec on the DEC-20 for 6000 lines of WordStar input. WordStar is not a well-designed formatter, in that it does not consistently identify document features. In particular, double spacing actually appears verbatim (and is converted to single spacing by WSLTEX, except at paragraph boundaries), and centered text (as in chapter titles) is just centered without any identification. Consequently, an editing session is still necessary to take the output of WSLTEX and convert it into true LaTeX input. The important thing is that every special feature in the input WordStar document is preserved, something which is not possible if the 8-th bit is just lopped off. Help yourselves. Extensions, modifications, and bug reports are welcome. It's all public domain, so don't sell it! ------- ======== ------- ========